Dan here…another post by Steve
Steve Roth | November 16, 2015
In a recent post Tyler Cowen makes an admirable effort to lay out his overarching approach to thinking about macroeconomics, revealing the assumptions underlying his understanding of how economies work. (Even more salutary, this has prompted others to do likewise: Nick Rowe, Ryan Avent.)
Cowen’s first assertion:
In world history, 99% of all business cycles are real business cycles.
This may be true, but it is almost certainly immaterial to the operations of modern, financialized monetary economies. He acknowledges as much in his second assertion:
In the more recent segment of world history, a lot of cycles have been caused by negative nominal shocks. I consider the Christina and David Romer “shock identification” paper (pdf, and note the name order) to be one of the very best pieces of research in all of macroeconomics.
That paper, which revisits and revises Friedman and Schwartz’s Monetary History, is clearly foundational to Cowen’s understanding of how economies work, so it bears examination — in particular, its foundational assumptions. The Romers state one of those assumptions explicitly on page 134 (emphasis mine):
…an assumption that trend inflation by itself does not affect the dynamics of real output. We find this assumption reasonable: there appears to be no plausible channel other than policy through which trend inflation could cause large short-run output swings.
This will (or should) raise many eyebrows; it certainly did mine. Because: it completely ignores the effects of inflation on debt relationships.
It’s as if Irving Fisher and Hyman Minsky had never written.
Assuming “inflation” means roughly equivalent wage and price increases, at least over the medium/long term (yes, an iffy assumption given recent decades, but…), inflation increases nominal incomes without increasing nominal expenditures for existing debt service. (Yes, with some exceptions for inflation-indexed debt contracts.) Deflation, the reverse. Nominal debt-service expenditures are (very) sticky. Or described differently: inflation constitutes a massive ongoing transfer of real buying power from creditors to debtors — and again, deflation the reverse.
“No plausible channel”?
Excepting one passing and immaterial mention of government debt, the the words “debt” and “liability” do not appear in the Romer and Romer paper, and it has only two passing mentions of “assets.” It’s as if balance sheets did not exist — which in fact they do not in the national accounting constructs then existing, that the Romers, Friedman, Schwartz, and presumably Cowen today are using in their mental economic models and in the “narrative” approach to explaining economies that Friedman, Schwartz, and the Romers explicitly champion.
If you go further and allow that wages and prices can inflate at different rates (which you must, given recent decades), you have extremely large and changing differentials between price inflation, wage inflation, and (especially) asset-price inflation.
All of these inflation dynamics are assumed away, made invisible and immaterial, in Romer and Romer — hence largely, at least presumably so, in Cowen’s understanding of economies. It is explicitly assumed (hence concluded) that those dynamics have no “real” effect. As in Romer and Romer, the words “debt,” “liability,” and “asset” are absent from Cowen’s “macroeconomic framework.” (Though he does give a polite if content-free nod to Minsky in his ninth statement.)
This explains much, in my opinion, about Cowen’s — and many other mainstream economists’ — flawed understanding of how economies work.
Cross-posted at Asymptosis.